Page:VCH Herefordshire 1.djvu/229

 ROMANO-BRITISH HEREFORDSHIRE history no less than its position.' ° Steep wooded hills, narrow and secluded valleys, mountain streams with swift currents, have effectually prevented any great possibility of a residential and peaceful occupation except in the two districts previously mentioned, round Hereford and Ross. Hence we find in those parts the greater proportion of the Roman remains, which, taken as a whole, are extremely scanty. The frequency of camps is the only feature which 'contradicts the barren- ness of the archaeological record. Herefordshire is a land of camps,'^ a fact which at first sight points to a lack of the signs of peaceful occupation as a probable feature. But out of a possible number of seventy earthworks in this small county, 30 miles in width and 40 miles in length, only one can certainly be classed as Roman. A Roman origin has indeed been claimed for others, and it is also suggested that the Roman occupation of the already existing British camps was of frequent occurrence.^ This is however unlikely, because the British camps were for the most part formed in sites which would not meet with favour, judged from the point of view of the Roman system of military tactics. For these reasons it is better to consider the site of Leint- wardine in the north of the county as that of the one authentically Roman camp.' On the whole, therefore, we are unable to regard Herefordshire as coming within the limits of the military district of Roman Britain. Lying as it does on the border of the hilly and unsettled region, it yet comes almost wholly within the area of the more peaceful ' lowland ' district of the midland plain. It comprises no Melandra or Gellygaer, and the one camp where the Romans defended themselves against the highland foes of the tribe of the Ordovices is of no great size or importance. There are no traces of troops or permanent garrisons, of civil administration or municipal life. The inhabitants were not Romans, but Romanized Britons.^" We are therefore prepared to find the characteristics of the Roman period similar in the main to those of the neighbouring counties of Shropshire and Worcester : a sparse and moderately- civilized population, Romanized, it is true, in language, arts and industries, yet in the absence of any large centre of life such as Glevum or Viroconium, leaving behind it but few traces of the life of the period. Its civilization was, as Pro- fessor Haverfield phrases it, ' Romanisation on a low scale.' Much that the same writer has said of its northern neighbour will apply equally well to our county. In the Middle Ages it was a border county covered with strongholds large or small, and houses fortified against the dangers of Welsh raid and pillage. In our own days it comprises . . . important agricultural areas, and most of it is well . . . inhabited. The Roman age offers neither border fortresses nor crowded populations. . . The district soon ceased to be a border fronting hostile tribes, and we may confidently include it among the peaceful portions of Roman Britain. On the other hand its civilian life did not develop widely ; on the contrary ... it was thinly peopled. . . The land lay remote from the centres of Romano-British activity, and still remoter from the centres of Roman Imperial life. Nor had it in itself any very obvious attractions for inhabitants.^^ Thus the district remained for nearly four centuries. We have no exact record of the time when the Romans were forced to abandon this part of ^ Arch. Sure. ofHerefs. (1896), 2. ' Ibid. 3 ; see generally art. 'Earthworks.' ' See for instance Watkin in ^rf^. Jcarw. xxxiv, 366 ; also Woolhope Club Trans. 1885, p. 335, where ' overwhelming reasons' are given for the Roman origin of Risbury Camp in Humber parish. ' See below, p. 1 84. '» Arch. Surv, 4. " F.C.H. Shrops. i, 2 1 5. I 169 22